Schizophrenia is a leading cause of disability striking relatively early in life, incurring high medical and mental health service costs. The emergence of atypical antipsychotic medications has offered patients greater opportunity to regain their functional status in society. However, the likelihood of sustained reduction in psychotic symptoms and prevention of relapse are largely dependent on treatment adherence with prescribed medications. Early treatment discontinuation not only renders treatment itself less effective, it also poses a major challenge in evaluating long term trial of effectiveness. In the landmark Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE), as in real world practice, participants often discontinue medication whether for lack of efficacy, side effects, social stigma, or subtler kinds of dysphoria. Reasons of discontinuation vary between different antipsychotic medications. Time to discontinuation was usually the primary study outcome because it was thought to reflect multiple factors that impede effective treatment (Kemmler et al., 2005;Lieberman et al., 2005), however, we need to fully understand the association between prescribed antipsychotic medications and different reasons of discontinuation and to carefully assess patient outcomes in longitudinal efficacy and cost measures (such as psychotic symptom severity, quality of life, drug and health care costs) resulting from the differences in treatment adherence that can be disrupted by competing causes of discontinuation. Furthermore, participants often miss one or more scheduled measurements before they even discontinue their treatment in longitudinal antipsychotic trials, which may also be related to the effectiveness itself and further complicates the interpretation of study results. This R21 is the first one that will examine the impact of both study discontinuation and intermittently missed measurements on the evaluation of cost/effectiveness in practical antipsychotic trials. It is also the first of which we are aware, that would simultaneously address the issues of effectiveness as assessed by the above longitudinal measures and cause- specific treatment discontinuation. The CATIE trial was the largest, longest study ever funded by NIMH but its findings have been contested because high attrition rate and intermittently missed measurements in the above outcomes have precluded reaching consensus on study conclusions. It is widely believed that the missing data results from and represents the unmeasured re-emergence of symptoms and corresponding increases in resource utilization and cost and thus can deeply distort study findings. This R21 is highly relevant to both in the evaluation of antipsychotic medications and to the development of methodological literature on the analysis of large longitudinal trials.